I’ve been working in software development and leadership for more than 20 years and I can tell you that good sofware requires good engineers. I’ve probably reviewed thousands of code files and worked with more than a hundred software developers over the years and if I have choice I always try to avoid hiring cowboys, hackers or “mad scientist” personalities. I prefer excellent, creative, dedicated engineers.

Software is not art, products are art, software is good craftmenship and engineering.

Good engineers are quite the opposite of hackers. They need to be able to write clean, well thought, documented and well architectured code, not to “hack together” something. The easy way is to hack something up over the weekend. But would the other team members be able to understand that code? will it scale? perform? be secure? will it be bug-free? can it easily be fixed? Good code is code that survives the test of time, not code that does something fancy, fast, and then needs to be thrown away.

Development speed is a measure not to be understimated, but the quality of the resulting code is not of lesser importance. It’s infinitely more challenging to build a successful company based on low quality code. Hiring a developer that is very talented and works very fast is useless if you have to throw away all his code and start over two months later. Moreover, the basis of your software codebase is going to set the tone for years to come. If the basis is bad or was done in rush, you’ll pay the price for a long time.

This makes me wonder what kind of software comes out from today’s popular hackathons. I guess that the better developers, later refactor and improve the code they wrote during the hackathon. That’s fine and reasonable, but not everyone can do that sucesfully, has the time and resources to do it or is willing to go through the pain and actually do it. Which means that many of the nowadays “hacker culture” companies end up with having low quality codebases and huge “maintenance” and “refactoring” expenses down the road. In some cases this even leads to failed products and companies.

Be a hacker entrepreneur, an inspiring product person, or a software engineer. Not the other way around.

“It seems that about 70% of users have registered the old fashioned way, without Facebook at all, so putting it only on Facebook would have closed it up for many users. Furthermore, only about 3% invited friends using the many mechanisms and incentives that we built! But some seem to have copy and pasted and sent the link by e-mail. I have a disturbing feeling that Facebook virality is mostly hype and that 90% of user never share a thing. The “social sharing” stuff is something everybody talks about but not many actually do it. That’s my impression of late. Another theory is that the “sharing” phenomenon slows down with time as people become indifferent and tired. What do you think?”

Over the years, I’ve worked with several companies doing the social dance. Several Facebook applications companies (dating, gaming), a social book reading web site, a social ratings service, helped with a few Facebook pages… etc.

This brings me to my feelings towards social sharing (trying to get people to virally spread your content):
1. Everybody is talking about social sharing, very few actually share content they discover.
2. Only certain, very narrow demographics/personalities share. Many shy away, don’t have time or simply don’t care.
3. As time passes, more people become indifferent or tired, and share less. I think this is the direction in which the market is going. I would expect less people to share as time passes.
4. Young adults with plenty of time on their hands seem to share more than older people. Curiously, children seem to share less than young adults.
5. Some people think that making things social and sharable is the answer to everything. Social sharing is not for every business and company. Not everything should have social features. Some things are naturally better when social, others are not.

Anyway, he kept asking the guy: Who would be your first adopters? Who’ll pay first for your product?
Those were great questions. That got me to remember a lesson I learned years ago, but one that I need be reminded of from time to time.

Back in 1996 I started my first start-up company. It was called “The Software Mint”. Actually I founded a computer games company several years before that, but that’s another story. My company was developing technology for building Virtual Worlds. This was back in the days before the first virtual world like games came out: Ultima Online, Meridian 59 and later EverQuest. It was the time when VRML (Virtual Reality Modeling Language) was getting started and way before Second Life.

Me and my co-founder developed a working prototype of a multi-user sync server, an objects data store and a graphical client software. It was all working pretty nicely, good enough for a demo with multiple users. I wrote a document describing the system and on I went to raise money for turning this into a real company. Needles to say I was young (22) and naive and knew nothing about raising investment.

My basic pitch was:
We are building this awesome virtual worlds enabling technology… What is it good for you ask?
Well, at first we’ll use it to build graphical chat rooms. As the technology matures, we’ll build virtual graphical shopping malls and stores, and eventually we’ll build sophisticated massively multi-player online games (MMOs or MMROPGs).

I didn’t see it back then but in retrospect, besides the market not being ready, the product ideas not being very good and my personal lack of experince… the main feedabck I was getting over and over again was: THE COMPANY LACKS FOCUS. The investors I met with were telling me that I should attack one market with one product and not try to build three products for three different markets. Obviously they were right, but I wasn’t smart enough to listen back then. Looking back this was probably one of my first hard learned lesson as an entrepreneur.

Why focus matters so much:
1. A start-up has limited resources, focusing more firepower on one problem or one market makes more sense than spreading thin.
2. Knowing your target market makes you sharp. You need to be able to imagine your client and his needs. If you can’t do that, you’re shooting in the dark.
3. Your company needs to get to market as soon as posisble. Focusing on one narrow segment or vertical will let you cut features and focus only on the things that mater most.
4. You need to prove your product idea has traction in order to succesfully raise investment. Thus focusing on the most likely to pay potential clients, makes the most sense.
5. Looking at history, companies trying to be all things for all people usually fail.

Unfortunately I still see companies lacking focus all the time. Just recently I was approached by a company building an “all purpose” security anti-piracy mechanism, looking for my help. They want to protect online music, games, software, CDs and what not. Reading their manifest can easily make one dizzy. If they would have picked one domain where they can do the most good and where they will find their most likely to pay clients, the’d fare much better than when trying to be all things for all people.

Abstract: Facebook games are extremely successful, because they are actually “casual” games with built-in viral distribution mechanisms, and not because they are really “social”. You are playing near your friends, not with your friends.

I’ve been playing computer games since 1986 and I’ve been occasionally developing games since 1991. I’ve played almost every conceivable genre and I’ve worked on many games and gaming related projects, including 2D, 3D, isometric, single-player, multi-player, casual, MMORPGs, Facebook… you name it.

In recent years, like everyone else, I’ve been playing various “social” games on Facebook. Games like Farmville, CityVille and Mafia Wars by Zynga, Happy Island by CrowdStar, Country Story and Pet Society by Playfish… and others. Most of these games are fun, albeit a bit too simplistic.  At some point I realized that these games are not at all social. If I have to distill it into one sentence: the social interaction in these games is mostly limited to begging your friends to click.

Let’s analyze the types of “social” behavior these games support and encourage:

1. Wall posts (sharing). Purpose and use:

- Asking friends to help you out by clicking.

- Bragging about in game accomplishments.

2. Sending Facebook Requests. Purpose and use:

- Asking friends to help you out by clicking.

- Inviting friends to play.

3. Comparing scores and achievements with your friends. Purpose and use:

- Bragging about in game accomplishments.

4. Visiting the game area of your friends (without ever meeting them). Purpose and use:

- Some kind of forced sense of togetherness.

Do these interactions warrant the title “social games”?

I don’t think so.

What kind of interactions will warrant this title?

I think that for a game to be called “social” it needs to support true “face to face” interactions with other human users. This should include seeing other players (or avatars) and communicating in person. All of these must be accomplished not through time shifted Facebook requests and wall posts, but rather in real time.

But wait… such games do exist…

It all started with Meridian 59, EverQuest and Ultima Online, then more recently with the likes of World of Warcraft, Habo Hotel, RuneScape, Dofus, Aion, Lineage and many others. I’d say these massively multiplayer on-line games (MMO / MMORPG) are the real “social games”. In these games users actually meet and interact with other users and not just send canned messages to one another.

Granted that many of these bug reports are so old that they are not relevant any more, but they are old because they haven’t been addressed in a timely fashion. 4500 open bugs is a huge number by any measure, even for Facebook. Originally I wanted to use the graph to estimate when the Facebook team will catch up and go over (not to mention fix) all of the bugs… unfortunately the current trend is upward. The answer at this point in time seems to be: never. I can only hope that the team dealing with this will be beefed up and they’ll manage to get a handle on this.

This might come as a shock, but perfectionism can hinder good software development, not to mention deadlines. Over time I’ve come to the sad realization that nothing in this world is ever perfect and the attempt to make it perfect is a never ending drain of resources.

Trying to get everything perfect, fixing all minor bugs that a user is not likely to encounter in 100 years or adding all newly requested features to the work scope, would most certainly lead your project to go over time and over budget. And more importantly it will considerably decrease the project’s chances of completion and success.

The trick is to settle for “good” for the upcoming release and to focus on “perfect” for the long run. A few tips:

• Correctly make the distinction between issues that are bugs and those that are new feature ideas/requests.
• Prioritize bugs and features and strictly act according to those priorities.
• Manage new feature ideas and requests, and be able to turn some of them down.
• Decide what bugs get fixed and what features get implemented according to your schedule and budget and not vice versa.

Dealing with Features

Each product or service can sport an endless number of features. A toaster can also make coffee and a video site can also support MP3 uploads, but should they? It’s very easy to get carried away and add unneeded features to any software. The more people you consult with the more feature suggestions you’ll collect and the longer the project goes on, more features will pile up on your To-Do list.
The trick is to have focus regarding what your product is and is supposed cater for, and to be able to stick to that. Many preach for you to come up with the minimum set of features you can get away with for your first version. I’d rather focus on a reasonable set of features and not on the minimum set. I think a product should be more than barely functional, but on the other hand it’s crucial not to get carried away and to know how to postpone features to the next versions.

Dealing with Bugs

Any seasoned practitioner of the software development art can tell you that there is no such thing as 100% bug free software. You might think it is bug free, but that’s only because you didn’t find all the bugs yet. This is true due the inherent complexity of most software projects and products. If you take any commercial software package or web application, given a bit of patience, you can find bugs in it. I’m aware of a few bugs lurking even in the word processor that I use to write these very lines.
The trick is to categorize and prioritize bugs and deal with the most important ones first in order to keep your milestones and launch your product or service on time (even if it still has some known bugs).

Dealing with the User Experience

Anything can be improved. There is practically no end to the number of improvements that can be made to any software, service or product in order to make the user happier. It’s a continuous cycle of designing, implementing, testing, and releasing new versions.
The trick is to start with something that is conceivably good. Then to solicit and collect user feedback, as well as use various user tracking tools to analyze where things go bad, and improve in the next version.

Side Note

At your next job interview a future employer might ask you: “What is your worst quality?”
It’s common wisdom that the least damaging answer is “I’m a perfectionist”, or is it?

Using this tip and a few others we achieved a 20% (!!!) conversion rate (registered to paying users) on this song licensing service.

Start-ups sometimes tend to be caught up with building up buzz, getting traction and with churning out features that they forget one of the most important facts of life:

The user is the center of your start-up, it’s all about the user and it will always be about the user. Pleasing the user has many facets; one of them is to provide more than adequate customer support. I’m going to risk and state the obvious, but you’d be amazed how many companies do not follow these simple rules:

1. You should answer to user e-mails and you should do it within 24 hours. I can’t emphasize this enough, this is the basis for everything else.

2. If you do not have an adequate answer, reply nonetheless and tell the user when his problem will be addressed/fixed/answered.

3. If you can budget a support team that’s great, if not, do it yourself. Yes, you, the CEO/CTO/Founder… reply to users personally.

4. Explain things to users, be transparent. Do not assume users are stupid.

5. Never blame the user. 99% of the time it’s your fault.

6. Take the opportunity to request feedback from the user. It would help you later.

7. If your user base is too big to handle, setup mass customer support means, like a forum, or user automated feedback tools. Monitor the forum and reply to posts. Assign moderators and advanced users to help you out with answering the newbie questions.

8. Provide one on one customer support to paying customers regardless of your user base size. If you don’t, they’ll soon not be paying customers.

9. If you are hiring a support team make sure they are properly trained, have good English and are courteous. Monitor their responses and make sure they do not do more damage than good. I’ve seen this happen many times.

10. Last but not least: Be personal. Users respond better to a flawed human that talks at eye level, than a heartless customer service machine.

The better you’ll do these things, the more users will think positively towards your company which in turn will translate into better business results.

If you got this far, please subscribe and get more tips and insight.

A Beginning

I’ve been planning to start a blog for over a year now, but having too much work and giving this a low priority had its toll. Finally I found the time to go ahead and do it, and here we are… you reading this, me planning my next posts…

A sense of purpose

As a freelance CTO (Chief Technology Officer), I work with many start-up companies and entrepreneurs, spanning many domains and industries. I get to examine and learn interesting new stuff all the time and I’d like to share. I plan to write analysis, tips, insights and pitfalls from my day to day experiences. It should be fun.

I have a lot to say, and if you are interested in my thoughts and knowledge, I suggest you click the subscribe button… thanks!

A short summary of my life

By the age of 17 I had started my own computer games development company. I’ve served as a software developer in the Israeli Military (a Mamram programmer in the Navy, to be exact). I completed my Computer Science degree on weekends. I worked for several tech companies, including IBM Research and a couple of start-ups, starting as a software developer and ending as VP R&D. I got married, have three daughters, and two cats. In recent years I’ve been working as a freelancer/consultant, mostly in the role of a hired gun CTO. I had the pleasure of working with more than a dozen start-up companies over the years. There are very few domains, technologies, programming languages or platforms that I did not happen to touch in some way or another (I dare you to challenge me!). I love learning and I make it habit to learn several new things each week. For a more complete overview, check out my LinkedIN profile or my (not up to date updated) web site, or my company.

Thanks for getting this far down the page. See you on my next post.